Three views, one process: reading this year’s mid-year outlook through The Plan My Wealth Way

What problem is a market outlook actually trying to solve?

Before we take any outside research seriously, we ask what problem it's actually trying to solve. Global X's mid-year piece is wrestling with three connected questions that have been dominating institutional conversations this year: is the AI spending cycle still early or maturing, has the reward for taking equity risk in Australia shrunk too far relative to income alternatives, and does the market need a broader hedge given a bumpy first half dominated by the Iran conflict and shifting Fed commentary?

Framed that way, it's less a set of predictions and more a map of the genuine open questions professional investors are debating right now. That's useful — but a map isn't a destination.

Reference: Global X 2026 Mid-Year Outlook (confidential), section headings "Where AI Meets Strategic Capital," "Strategic Positioning for Income, Value-Disciplined Growth and a Broadening Rally," and "Broadening Beyond Equities: The Case for Commodities in H2." The framing of these as three connected "questions" is Plan My Wealth's own synthesis, not Global X's wording.

What does the evidence actually show?

Stripped back to the underlying data points (rounded, and read alongside other sources rather than taken in isolation), a few things stand out:

• AI spending, measured against the size of the global economy, still sits well below where past build-outs like mobile networks, the internet and the railways peaked (each of those reached roughly two to three times AI's current share of global GDP at their busiest) — though monetisation, in the form of cloud and frontier-model revenue growth, has been accelerating quickly alongside the spending, which is the more recent and arguably more important development.
• The nature of AI compute demand is shifting. A growing share of the industry's spending is now going toward running existing models (inference) rather than training new ones, and that share is expected to keep rising over the next couple of years — which matters because inference is an ongoing, recurring cost rather than a one-off build, tightening demand for memory chips and related hardware on a sustained basis rather than in bursts.
• Value creation has so far flowed to chipmakers, then to the cloud giants themselves — each of those two groups has added trillions of dollars in market value over the past few years — with the physical infrastructure layer (power, grid and data-centre enablers) not yet having had the same re-rating despite hyperscaler capex now running in the hundreds of billions annually and trending toward roughly US$1 trillion a year by the end of the decade.
• Australia's equity risk premium has fallen to roughly a 25-year low, and local dividend yields are sitting near their lowest levels since the early 2000s (excluding the COVID period, at somewhere around 3.5% against a 20-year average closer to 4.4%, as at 30 June 2026) — a genuine structural shift in the reward for holding growth assets versus income alternatives.
• Small and mid-cap companies are trading at a rare valuation discount to large caps (as at 17 June 2026) — among the cheapest relative valuations seen in close to two decades — after years of the opposite being true, alongside improving fundamentals on measures like return on equity and free cash flow yield.
• Energy storage demand has three separate tailwinds converging at once — renewed growth in electric vehicle adoption, data centres increasingly needing on-site battery storage simply to get connected to a constrained grid, and broader clean-energy storage build-out — with AI datacentre power needs reportedly growing several times faster than grid capacity itself as at early-to-mid 2026, leaving storage as one of the few practical ways to bridge that gap in the near term.

A note on timing: the figures above reflect Global X's research as at various points in June and July 2026, ahead of this article's 6 August 2026 publication date. Markets move; treat these as a snapshot of the evidence at the time, not a live feed. Check the date on any figure before relying on it.

References for this section: Global X 2026 Mid-Year Outlook (confidential; figures rounded and paraphrased), which in turn cites IMF, GSMA, ITU, OECD, UNCTAD and Maddison Project historical datasets (AI spend vs GDP); Barclays Research, Morgan Stanley Research and Deloitte estimates (inference vs training compute share); company data and Global X ETFs analysis (hyperscaler capex and AI infrastructure value creation); Bloomberg data as of mid-2026 (Australian equity risk premium and dividend yields); Bloomberg and MSCI data (SMID valuations); and the IEA's Energy and AI research (datacentre power demand vs grid capacity).

Each of those points is defensible evidence. None of them, on its own, tells you what to do with your portfolio.

What do the numbers actually show once you run them through?

This is where a research view has to earn its keep. It's one thing to say "income strategies look more attractive given where dividend yields and bond yields now sit" — it's another to actually model what that means after tax, after volatility, and against your specific drawdown needs.

Global X's own analysis, for instance, models how recent CGT settings shift the after-tax comparison between an income-tilted approach and a pure growth approach over a multi-year holding period — and the gap, once tax is accounted for, is smaller than the pre-tax numbers alone would suggest. That's a useful directional insight. It's not a substitute for running the same maths against your actual balance, your actual tax position and your actual timeframe, which is the step that turns a general observation into something specific to you.

Reference: Global X 2026 Mid-Year Outlook (confidential; hypothetical worked example, not client-specific), using Bloomberg data and standard Australian marginal tax and CGT assumptions consistent with ATO settings and the RBA's inflation target band. Paraphrased and rounded here, not reproduced.

What happens under each of the three scenarios?

Because we're not required to pick a single one of the three lenses as "correct," the more useful exercise is scenario testing: if the Bull is right and AI infrastructure spending keeps broadening, what happens to a well-diversified portfolio? If the Strategist is right and markets rotate toward small caps and value, does the same portfolio still hold up? If the Cautious view proves closer to the mark and commodities or gold need to play a bigger defensive role, has the portfolio got room to absorb that without a wholesale rebuild?

A portfolio that only performs well under one of those three outcomes isn't diversified — it's a bet. The modelling step is where we check for that before any recommendation is made, not after.

Source: no external citation — this scenario-testing approach is Plan My Wealth's own methodology, applied to the three lenses set out in the Global X 2026 Mid-Year Outlook (confidential).

What do terms like GARP, SMID and ERP actually mean?

A lot of this research is written for other professionals, which means it's dense with shorthand. In plain English:

Equity risk premium

The extra return shares are expected to deliver over safer assets like bonds, to compensate for the extra risk. Shrinking, as it has in Australia, means less reward for the same risk.

GARP

Growth at a reasonable price. An investing style that looks for genuine growth companies without paying the very highest prices for it — a middle path.

SMID

Small and mid-cap companies — smaller than the well-known household names often used as shorthand for "the market."

De-dollarisation

Central banks gradually diversifying their reserves away from the US dollar, often by holding more gold instead.

Understanding these terms isn't about being able to use them at a dinner party — it's so that when we do bring a recommendation to you, you're evaluating the reasoning, not just trusting the label.

Source: standard financial terminology, independent of Global X; definitions are Plan My Wealth's own plain-English explanations, not quoted from any single source.

What does this actually mean for your portfolio?

Here's the honest summary. This outlook doesn't tell any PMW client to buy more AI infrastructure exposure, rotate into small caps, or add gold. What it does is confirm that the range of plausible outcomes over the next six to twelve months is genuinely wide — wider than usual — and that a portfolio built to survive all three of these lenses, rather than bet on one, is the more defensible starting point.

For clients at or near retirement in particular, that usually means: a diversified growth allocation that isn't overly concentrated in any single theme (AI included), an income component that reflects today's genuinely different yield environment rather than last decade's, and enough of a defensive buffer that none of the three scenarios above — bullish, rotational or cautious — forces a bad decision at the wrong time. If you'd like a hand structuring a portfolio built around your own situation rather than a template, or just want to see how your current mix holds up against these three lenses specifically, that's exactly the kind of review worth booking in.

Source: no external citation — this section is Plan My Wealth's own general commentary and view, not attributed to Global X or any third party.